Ask people involved in a church plant that closed what went wrong and you will usually be told about the year it collapsed. Funding dried up. The core team moved away. The planter burned out. A conflict split what was left.
Ask them what they thought on the day it launched and a different answer often appears. Several people had reservations. Nobody said them out loud, because the date was set.
Why the date gets set first
A launch date is enormously useful. It focuses effort, it gives supporters something to give towards, it makes the thing real. It is also, almost always, chosen before anybody knows whether the plant is ready — because readiness is hard to assess and a date is easy to announce.
Once announced, the date acquires gravity. Money has been raised against it. People have made arrangements. A venue is booked. Announcing a delay now means explaining to everyone who committed that you were wrong, and most people would rather launch a plant they doubt than have that conversation.
So the date stops being a plan and becomes a constraint. Everything else adjusts around it, including the assessment of whether to proceed.
Five things that were usually missing
Across accounts of plants that struggled, the same gaps recur. None of them is dramatic.
No one who could say no. The planter was sent by a church that celebrated but did not oversee, or by a network that funded but did not govern. Nobody had both the standing and the information to say: not yet.
A core team assembled by enthusiasm. Twenty people who liked the vision, rather than eight who had agreed what they were committing to, for how long, and what they would do when it was difficult.
No financial floor. A budget built on optimistic giving from people who had not yet been asked, with no honest answer to what happens in month fourteen if it does not arrive.
Safeguarding deferred. A policy to be written once things settle down, in a congregation that from day one included children and vulnerable adults.
Nothing agreed about failure. No definition of what "not working" would look like, and no agreement about who decides, so the only way to stop was for someone to break.
The awkward question
The most useful question a sending church can ask is not "is this planter called?" It is: if this does not work, how will we know, who decides, and what happens to the people?
That question is unpopular because it feels like doubt. It is the opposite. Being willing to name a failure condition in advance is what makes it possible to try something risky without gambling with other people's lives. A planter who cannot answer it has not thought about the congregation's exposure, only their own conviction.
The second question is quieter: who can tell you no, and have you actually given them the information they would need? Oversight that only sees the reports the planter chooses to send is not oversight.
What honest support looks like
If you are supporting a plant, the most valuable thing you can offer is often the least welcome. Slow the date. Fund the governance before the launch event. Insist that safeguarding exists before the first gathering rather than after the first incident. Ask for the budget's pessimistic case, and ask what happens in it.
And be willing to fund a plant that does not launch this year. A supporter who will only pay for momentum is part of the mechanism that sets dates badly.
When one closes
Some plants close, and that is not a scandal. Ministry involves attempting things that might not work, and an environment where nothing is allowed to fail is an environment where nothing honest gets reported either.
The relevant questions when one closes are narrow and answerable. Were the people cared for? Was anyone financially exposed who should not have been? Was anything learned, and written down where the next planter can find it? Did anyone tell the truth about it publicly, or did it quietly disappear from the count?
That last one matters more than it sounds. An organisation that reports its plants and never reports its closures is not reporting. It is advertising.